This study applies artificial neural networks (ANN) to simulate the soil stress-strain relationship observed in test data from triaxial shear testing of saturated sand. A genetic algorithm is used to obtain an optimal framework for the ANN. The results show that the proposed genetic adaptive neural network (GANN) can effectively model undrained monotonic and cyclic triaxial behaviour of saturat...

A granular material is most likely to possess an initial anisotropic structure due to the depositional process or the previous stress history. There is also induced fabric anisotropy during shearing. Concerning the significant influence of fabric anisotropy on the stress-strain behaviour of a granular material, in-depth investigation on the two kinds of fabric anisotropy is essential. Initial f...

Triaxial compression and extension tests have been conducted under different initial anisotropy conditions to investigate the undrained response of a crushed silica sand. The loose medium specimens were prepared using moist tamping method. Five stress paths with ratios (q/p?) employed prepare anisotropically consolidated specimens. Several specific condition in which rotation occurred monotonic...

The principal objective of this thesis is to assess the capabilities of an already established analytical framework for understanding and predicting the behavior of piles driven in highly overconsolidated clays (OCR24). The skin friction mobilized during pile loading to failure depends on the sequence of preceding events, i.e. pile installation and soil consolidation. This research utilizes the...